QtKat - LyCORIS LoCon - PonyDiffusionXLV6

LoCon trained on qtKat's artstyle for ponydiff: https://civitai.com/models/257749

Artist: Twitter gets banned alot, look him up on Danbooru for his socials

Because of my lack of experience and time, the version number may be based on changes that have made certain aspects worse. As such a previous version may perform better than a later version.

I'm still learning the ins and outs of the model and LoRA training in general, please leave reproducible feedback or tips so that I may improve over time.

Check the "About this version" section for any quirks or pitfalls I'm aware of or have experienced.

Some Q's:

1. What's your workflow?

   • Sampler from images is done with Restarts with segments of [3,2,0.06,0.30], which are values from the paper on Restarts

   • Initial resolution is chosen from appendix I in the paper on SDXL

     • Source: https://arxiv.org/abs/2307.01952

   • Upscale is by 1.5x

   • Upscaler is usually 4x_fatal_Anime_500000_G or 4x_NMKD-Siax_200k

   • Second sample pass is DDIM/DDIM_Uniform at .5 denoise 30 steps w/ no restarts

   • V10 Second sample pass is Euler/normal at .4 Denoise 20 steps w/ restarts

   • A facedetailer is used

   • I use ComfyUI, sorry

2. What strength should I be using?

   • 1, but if it's too much .8 is very good as well

3. Pairs of images?

   • Left image is base Pony, Right image is Autismmix

   • V10 All images are Autismmix Confetti

4. What does your filename structure mean?

   • It's split into 4 parts of [XX]-MAnon-[YYYY]-[ZZZZZZ]

     • [XX] - Artist initials

     • MAnon - My moniker

     • [YYYY] - SD version and Model version (Ex. XLV4.1 is on SDXL [Not 1.5] and it's version 4.1 of the model)

     • [ZZZZZZ] - The epochs, only useful to me